Vascular Perfusion and Blood-Brain Glucose Transport in Acute and Chronic Hyperglycemia

Journal of Neurochemistry Raven Press, Ltd., New York 0 1988 International Society for Neurochemistry

Vascular Perfusion and Blood-Brain Glucose Transport in Acute and Chronic Hyperglycemia Sami I. Harik and Joseph C. LaManna Department of Neurology, University Hospitals of Cleveland and Departments of Neurology, Pharmacology and Physiology, Case Western Reserve University School of Medicine, Cleveland, Ohio, U.S.A.

Abstract: We studied the effects of acute and streptozotocininduced chronic hyperglycemia on regional brain blood flow and perfusion characteristics, and on the regional transport of glucose across the blood-brain barrier in awake rats. We found ( I ) a generalized decrease in regional brain blood flow in both acute and chronic hyperglycemia; (2) that chronic, but not acute, hyperglycemia is associated with a marked and diffuse decrease in brain L-glucose space; and (3) that chronic hyperglycemia does not alter blood-to-brain glucose transport. Taken together, these results suggest that in strep

tozotocin-induced chronic hyperglycemia,there is a reduction in the proportion of perfused brain capillaries and/or an alteration in brain endothelial membrane properties resulting in decreased noncarrier diffusion of glucose. Key Words: Streptozotocin-induced diabetes-Regional brain blood flow-Glucose transport-Blood-brain barrier-L-Glucose space-Brain capillary recruitment. Harik S. I. and LaManna J. C. Vascular perfusion and blood-brain glucose transport in acute and chronic hyperglycemia. J. Neurochem. 51, 1924-1929 (1988).

Recent evidence suggests that hyperglycemia has deleterious effects on ischemic brain injury in experimental animals and humans (Meyers and Yamaguchi, 1977; Siemkowicz and Hansen, 1978; Ginsberg et al., 1980; Pulsinelli et al., 1982, 1983; Longstreth et al., 1983),possibly due to production of high levels of brain lactic acid and resultant tissue acidosis. Although diabetes mellitus, with its associated hyperglycemia, is considered a risk factor in clinical cerebrovascular disease (Riddle and Hart, 1982; Pulsinelli et al., 1983), little is known about either the effects of chronic hyperglycemia on regional cerebral blood flow (rCBF) and vascular perfusion characteristics or how the brain reacts to ischemic and hypoxic insults. Duckrow et al. (1985, 1987)reported decreased CBF during acute and chronic hyperglycemia in the rat and concluded that it is not likely due to a decrease in the cerebral metabolic rate (Duckrow and Bryan, 1987). Experimental evidence also suggests that sustained hyperglycemia in rat models of diabetes mellitus is associated with decreased blood-to-brain glucose transport (Gjedde and Crone, 1981; McCall et al., 1982), which is manifested by a lower calculated maximum

transport of glucose in rats with chronic hyperglycemia (Gjedde and Crone, 1981). This could be caused by “down-regulation” in the number of active glucose transporter molecules in endothelial membranes of brain capillaries, which constitute the blood-brain barrier (BBB). Such a decrease in the BBB capacity to transport glucose in chronic hyperglycemia could be interpreted as a mechanism to protect the brain from the effects of increased tissue glucose and glycogen, and might explain clinical observations that suggest that brain glucose is deficient in diabetic subjects who undergo rapid normalization of their plasma glucose levels (DeFronzo et al., 1980). The purpose of this study was to investigate the effects of acute and chronic hyperglycemia on rCBF and perfusion characteristics, and on the unidirectional BBB transport of glucose in awake rats. Our results indicate that both acute and chronic hyperglycemia decrease CBF, and that chronic but not acute hyperglycemia is associated with a decrease in the brain Lglucose space, a finding that may indicate a reduced number of perfused capillaries. However, our findings do not support earlier suggestions that blood-to-brain

Received April 15, 1988; revised manuscript received June 28, 1988; accepted July 8, 1988. Address correspondenceand reprint requests to Dr. S. I. Hank at Department of Neurology, University Hospitals, Cleveland, OH 44106. U.S.A.

Abbreviations used BBB, blood-brain banieq CBF, cerebral blood flow; PS, permeability-surf-ace area; CBF, regional cerebral blood flow.

1924

CBF AND BBB GLUCOSE TRANSPORT IN HYPERGLYCEMIA glucose transport is decreased in chronic hyperglycemia (Gjedde and Crone, 1981; McCall et al., 1982). Part of this work was presented in abstract form (Hank et al., 1986; LaManna et al., 1987).

MATERIALS AND METHODS Experimental procedures Adult male Wistar rats (300-350 g) were used in all experiments. Three groups of hyperglycemic rats were used. In one group acute hyperglycemia was induced by the intraperitoneal injection of 2 ml of 50% glucose solution, about 20 min before the experimental measurements were made. In the other two groups chronic hyperglycemia was induced for 2 or 4-5 weeks by a single injection of streptozotocin (60 mg/kg, i.p.). The magnitude of hyperglycemia was similar in the acute and chronic groups (Table 1). The streptozotocintreated rats developed hyperglycemia (plasma glucose >20 mM) and failed to gain weight. Two and 4-5 weeks after the injection of streptozotocin, rats weighed about 70 and 100 g less than control rats, respectively. A fourth group of rats were left untreated and served as the normoglycemic control group. On the day of the experiment, rats were anesthetized with chloral hydrate (400 mglkg). Then cannulae were inserted into the tail artery and into the right atrium of the heart via the external jugular vein. The skin was infiltrated with local anesthetic solution, then sutured. The rats were restrained in padded plaster casts after they recovered from anesthesia. Experiments were performed 3-4 h after surgery, when the rats were fully awake. Body temperature was monitored and maintained at 37OC by a rectal thermister probe connected to an infrared heating lamp. The arterial cannula was used to sample blood and plasma for gases, pH, hematocrit, and glucose determinations. The physiological variables for all groups are listed in Table I. Regional blood-to-brain glucose transport, as well as CBF and perfusion characteristics, were determined by the double label, single pass, atrial bolus injection method (LaManna and Hank, 1985, 1986). The arterial cannula was connected to a syringe fitted to a withdrawal pump calibrated at 1.60 ml/min. The withdrawal pump was started and within seconds a 150 pl bolus of physiological buffer solution was injected into the right atrium. The buffer contained 10 pCi of n-[14C]butanol(1.0 Ci/mol; New England Nuclear) and 2550 pCi of D-[3H]glucose or ~-[~H]glucose, and an amount of unlabeled glucose to match the plasma glucose concentration. ~-[3,4-~H(N)]Glucose (48.9 Ci/mmol; New England Nuclear) and L-[ l-3H(N)]glucose (10.7 Ci/mmol; New England Nuclear) were evaporated to dryness under reduced pressure, just prior to use, to remove volatile tritiated contaminants. Ten seconds after the atrial bolus injection, the rats were decapitated and simultaneously the withdrawal pump was stopped and the arterial cannula removed. The withdrawn arterial blood was quantitatively transferred to tared vials, and aliquots were measured for radioisotope content. The brain was rapidly removed and bilateral samples from the frontal and parietal cerebral cortex, hippocampus, cerebellum, and striatum were weighed and their radioisotope content determined by beta scintillation spectroscopy. Samples of blood (100 pl) oozing from the foramen magnum were collected in heparinized tubes, and aliquots of the plasma were counted to estimate the radioactive content of the ce-

1925

rebral vascular compartment at decapitation (Sage et al., 1981). Plasma, rather than whole blood, was counted because neither D-glucose nor glucose readily penetrates rat erythrocyte membranes. On the other hand, the ['4C]butanol content of blood sampled from the severed head was negligible. Precautions were taken to minimize the evaporation of [ 14C]butanolfrom tissue and blood samples.

Calculations rCBF in milliliters per 100 g per minute was calculated by the butanol indicator-fractionation method (Sage et al., 198l), without correction for incomplete butanol extraction, as follows: rCBF = Fs X I4C brain/(14C, X brain weight) X 100 where F, is the withdrawal rate of the syringe (milliliters per minute); I4C brain is the radioactive content, in dpm, of the brain sample; l4CSis the radioactive content, in dpm, of the withdrawn blood and the brain weight is in grams. We did not correct for butanol efflux from the brain during the 10 s that elapsed between injection of tracer and decapitation, but our results underestimate rCBF by no more than 10%(Sage et al., 1981). To estimate specific carrier-mediated @glucose transport across the BBB, we had to account for contamination by residual intravascular tritium and any D-glucose that entered the brain by simple diffusion. To accomplish this, the volume of distribution of L-glucose (microliters per gram) was determined for each brain region in rats of each experimental group by dividing the radioactive content of brain samples (dpm per gram) by the concentration (dpm per microliter) of L-glucose in plasma of the venous blood oozing from the severed head. The concentration of plasma D-[3H]glucose in the venous return from the head was multiplied by the previously calculated mean regional glucose space to estimate the radioactivity in tissue samples that were contributed by intravascular tracer and non-carrier-mediated diffusion. Unidirectional blood-to-brain glucose influx ( J ) was calculated as described by Gjedde (1983): J = -PF X In (1 - E ) X G where PF is plasma flow (derived by multiplying CBF by the volume fraction of plasma in arterial blood); G is plasma glucose concentration; In is the natural logarithm and E is the extraction fraction of Dglucose, calculated by dividing the ratio of 3H to 14C in each brain sample by the ratio of 3H to 14Cin the withdrawal syringe (Pardridge and Oldendorf, 1975). The extraction fraction was corrected by subtracting the counts attributed to intravascular content and simple diffusion; the correction factor in the cerebral cortex amounted to -30% of the total counts at plasma glucose concentrations of about 30 mM. Values from the two sides of the same brain were averaged and the means used for data analysis. Statistical differences among the expenmental groups were computed by one-way analysis of variance, followed by Dunn's multiple comparison method (Kirk, 1982). Significance was considered at p < 0.05.

RESULTS Arterial blood gases and hematocrit were comparable in all experimental groups, but arterial blood pH was significantly lower in rats that were hyperglycemic for 4-5 weeks (Table 1). Mean plasma glucose in normoglycemic control rats was 10.1 mM. Mean plasma glucose concentrations were similar in the acute and f . Meuroehem., Vof.51. No. 6, 1988

S. I. H A N K AND J. C. LAMANNA

1926

TABLE 1. Physiological values in experimental groups of awake rats Hyperglycemia

p.0~(mm Hg) P.co2 (mm Hg) PH Hematocrit (%) Glucose (mM)

Normoglycemia

Acute

2 weeks

4-5 weeks

91 f 4 35 f 1 7.37 f 0.01 46 f 1 10.1 f 0.6"

98 f 3 36 f 1 7.35 f 0.01 48 f 1 32.3 f 2.1

96 f 2 37 f 1 7.36 f 0.01 41 f I 26.5 f 0.8

90 f 3 37 f 1 1.27 f 0.03" 46 f 1 28.8 f 1.6

Values denote means f SEM for 18 rats in the normoglycemic group, 2 1 rats in the acute hyperglycemic group, and 12 and 1 1 rats in the 2- and 4-5-week chronic hyperglycemic groups, respectively. " Significantly different from the other groups at p < 0.01.

chronic hyperglycemia groups: about threefold that of the control group (Table 1). Two weeks of streptozotocin-induced chronic hyperglycemia had no effect on the water content of the cerebral cortex (see Table 1 of the reference by Harik et al., 1988). Regional blood-to-brain glucose influx was similar in acute and chronic hyperglycemia except in the hippocampus where it was significantly lower in rats with 4-5 weeks of hyperglycemia (Table 2). Unidirectional carrier-mediated influx of D-glUCOSe in the hyperglycemic groups was about 80% of the maximal glucose transport that we previously reported in awake rats by the same method (LaManna and Harik, 1985). In these experiments, we determined unidirectional influx of D-glUCOSe at a single hyperglycemic plasma glucose concentration greater than five times the K,. There was a tendency for D-glucose transport to increase in all brain regions after 2 weeks of hyperglycemia, but this trend was reversed by 4-5 weeks of sustained hyperglycemia (Table 2). Except for reduced g glucose influx in the hippocampus of rats with 4-5 weeks of hyperglycemia, our results are not consistent with previous reports suggesting decreased blood-to-brain glucose transport in chronic hyperglycemia (Gjedde and Crone, 198 1; McCall et al., 1982). We have no expla-

TABLE 2. Blood-to-brain D-glucose influx in acute and chronic hvDeralvcemia Chronic hyperglycemia

Cerebral cortex Frontal Parietal Hippocampus Cerebellum Striatum

Acute hyperglycemia

2 weeks

4-5 weeks

216 f 15 213 f 12 156 f 9 196 f 1 1 181 & 17

240 f 23 235 k 19 169 f 10 222 f 15 189 f 14

220 f 19 204 f 16 110 f 15" 184 f 26 157 f 23

~~

~~

Carrier-mediated glucose transport values in pmol/ 100 g/min are the means f SEM for 13 rats in the acute hyperglycemic group and for nine and six rats in the 2- and 4-5-week chronic hyperglycemic groups, respectively. Significantly different from the other groups at p < 0.05.

J. Neurochem., Vol. 51, No. 6 , 1988

nation for the hippocampal results in the chronic hyperglycemia group. We also found a global decrease of up to 30% in CBF in acute hyperglycemia, which was statistically significant, in the frontal and parietal cerebral cortex (Table 3). Hyperglycemia for 2 weeks caused a further decrease in CBF in all the brain regions (Table 3). With continued hyperglycemia for 4-5 weeks rCBF values showed a trend toward recovery, but remained significantly less than those in the frontal and parietal cortex and cerebellum of normoglycemic rats (Table 3). We previously reported that acute hyperglycemia increases the brain L-glucose space in awake rats, probably because of increased passive diffusion of L-glucose (LaManna and Harik, 1985). In this study we also found a trend toward higher L-glucose space values in rats with acute hyperglycemia (Table 4). Decreased brain L-glucose space in rats subjected to streptozotocin-induced chronic hyperglycemia (Table 4) was an unexpected finding. DISCUSSION Our results do not agree with the conclusions of Gjedde and Crone (1981), who reported that 3 weeks of streptozotocin-inducedhyperglycemia caused a 30% decline in the maximum glucose transport capacity of the BBB. Methodological differences may explain this disparity as follows. (1) The previous study was conducted in lightly anesthetized rats, whereas we used awake rats; differences in rCBF and blood-to-brain transport properties between awake and anesthetized rats do exist (LaManna and Harik, 1986). (2) Gjedde and Crone estimated the vascular and diffusional components of glucose uptake from prior work and assumed they were the same in acute and chronic hyperglycemia. Our results clearly show that this is not the case in awake rats (Table 4). Taking into our calculations the differences in brain L-glucose space between acute and chronic hyperglycemic rats probably further explains the discrepancies between the two studies. The present data also do not support the conclusions of McCall et al. ( 1982), who found decreased

CBF AND BBB GLUCOSE TRANSPORT IN HYPERGLYCEMIA

1927

TABLE 3. Regional brain blood flow in acute and chronic hvDernlvcemia Hyperglycemia Acute

2 weeks

(20) 117 f 7” (20) 119 f 7“ (20) 93 f 6 (20) 73 f 3 (16) 116 -I- 10

(14) 99 +9’ (14) 104 + 8’ (14) 75 k 6 “ (14) 56 ?4’,‘ (14) 88 + 8

Normoglycemia Cerebral cortex Frontal Parietal Hippocampus Cerebellum Striatum

+

(14) 153 9 (14) 148 f 7 (14) 103 f 5 (14) 8 6 f 3 (10) 121 f 8

4-5 weeks (11) (11) (11) (11) (11)

107 f 8’ 113 -I- 7“ 95+9 69f5” 1 2 0 k 11

CBF values, in m1/100 g/min, are the means f SEM of the number of observations in parentheses. Significantly different from the normoglycemic group at ‘p < 0.05 and bp < 0.01. There were no significant differences in rCBF results among the hyperglycemic groups, except in the cerebellum, where the acute hyperglycemicgroup was significantlydifferent from the 2-week chronic hyperglycemic group at ‘p < 0.0 1.

extraction of glucose relative to water when both were injected into the carotid arteries of streptozotocintreated rats. Gjedde and Crone (1981) suggested “down-regulation’’ of glucose transporters in brain capillariesas a possible explanation for decreased maximum glucose blood-to-brain transport that they calculated in chronic hyperglycemia. However, our recent work indicates that the density of glucose transporters, measured by specific D-glucose-displaceable [3H]cytochalasinB binding, is increased, not decreased, in cerebral capillaries of rats with chronic streptozotocin-induced diabetes (Harik et al., 1988). On the other hand, the weight of recent evidence supports our contention that glucose transport across the BBB is not altered in chronic hyperglycemia. Duckrow et al. (1988) found no difference in the permeability-surface area (PS) product for Dglucose between rats with acute and chronic hyperglycemia. Pelligrino and Sharp (1988) measured blood-to-brain Dglucose influx in streptozotocin-treatedchronic hyperglycemic rats that were made acutely normoglycemic, and found no difference in glucose influx between these rats and controls. Also, positron emission tomography studies of regional cerebral glucose transport in diabetic

subjects suggested that both the number and affinity of D-glucose carriers in the BBB are within normal limits (Brooks et al., 1986). The results ofthese studies, which employed experimental methods and paradigms different from ours, further support our conclusions. We also found that hyperglycemia causes a significant drop in rCBF. Acute hyperglycemiainduces a drop of up to 30%in CBF when compared to that of awake normoglycemic rats. A further decrease of about 15% accompanies chronic hyperglycemia (Table 3). These findings are consistent with the results of Duckrow et al. (1985, 1987) and Ginsberg et al. (1980). Rubin and Bohlen (1985) reported a larger diameter of pial arterioles in streptozotocin-treated rats. Also, the velocity of erythrocytes through small vessels, and thus volumetric blood flow, was increased in these rats. However, because tissue Po, did not increase, Rubin and Bohlen concluded that although blood flow per capillary unit may have increased, the number of perfused capillaries must have decreased in their experimental rats. Finally, we found a major, statistically significant, drop in the brain L-glucose space of rats that are rendered chronically hyperglycemic, compared with those with acute hyperglycemia (Table 4). Regional differ-

TABLE 4. L-Glucose space in acute and chronic hyperglycemia Hyperglycemia

Cerebral cortex Frontal Parietal Hippocampus Cerebellum Striat um

Normoglycemia

Acute

Chronic

(8) 18.9 f 2.4 (8) 20.4 k 2.7 (8) 18.0 f 1.6 (8) 28.5 f 3.1 (4) 12.3 f 1.6

(8) 22.7 f 1.5 (8) 23.5 f 1.5 (7) 22.1 f 1.8 (8) 30.8 f 2.5 (5) 17.8 -I- 1.3

(10) 13.8 f 1.3” (10) 14.4 f 1.3” (10) 15.4 f 1.7’ (10) 19.5 f 1.7“.‘ (10) 12.5 f 1.5

L-Glucose space values in pl/g, are the means f SEM of the number of observations in parentheses. There were no differences in the L-glucose space values between rats with 2 or 4-5 weeks of hyperglycemia, and the results from both groups were combined. Significantly different from the normoglycemic group at ‘p < 0.05, and from the acute hypedycemic group at “p < 0.01, and at ’p < 0.05.

J. Neurochem., Vol. 51, No. 6 , 1988

1928

S. I . HARIK AND J. C. LAMANNA

ences in the L-glucose space in normoglycemia are consistent with our prior observations (LaManna and Hank, 1985, 1986). The brain L-glucose space reflects two components: trapped residual intravascular tracer, and tracer that crossed the BBB by non-camer-mediated diffusion. The important contribution of the first component is evidenced by similar regional variations that were noted with vascular markers like sucrose, inulin, and "Cr-tagged erythrocytes (Blasberg et al., 1983; Cremer and Seville, 1983; Cremer et al., 1983). The second component is directly proportional to plasma glucose concentration and is nonsaturable. The brain L-glucose space increases in acute hyperglycemia when compared to normoglycemia (LaManna and Hank, 1985). However, persistent hyperglycemia for 2-5 weeks results in a decline in the brain L-glucose space to values that are 30-40% lower than those of acute hyperglycemia. The results concerning the glucose space in chronic hyperglycemia reflect changes in the PS product for Lglucose. Because the method we used does not distinguish between the two components of the PS product, our results could be due to decreased permeability of the brain endothelium to L-glucose, decreased vascular surface area, or to a combination of these factors. Decreased permeability could be caused by a long-term effect of streptozotocin treatment on the diffusion of glucose across endothelial membranes. The differences in the L-glucose space could also be due to shrinkage of the perfused intravascular space in the brains of rats with chronic hyperglycemia. This, coupled with lower rCBF, suggests that the density of perfused brain capillaries is decreased in rats given streptozotocin. This phenomenon could be caused by a decrease in the fraction of perfused brain capillaries-decreased capillary "recruitment"-or by a diminished anatomic density of brain capillaries, as suggested by Jakobsen et al. (1987). The mechanisms by which streptozotocin treatment causes decreased regional L-glucose space and rCBF remain to be determined. Irrespective of mechanism, these results can have important pathophysiological consequences that may be related to the deleterious effects of hyperglycemia on clinical and experimental ischemic brain injury (Meyers and Yamaguchi, 1977; Siemkowicz and Hansen, 1978; Ginsberg et al., 1980; Pulsinelli et al., 1982, 1983; Longstreth et al., 1983). However, it should be pointed out here that streptozotocin induces a diabetic state with acidosis, a condition that may not be ascribed simply to hyperglycemia. In any case, because of the prevalence of diabetes mellitus in the general population, and because of the relationship between diabetes and the incidence of cerebral vascular diseases, vascular perfusion of the brain in animal models of diabetes is a subject that should be investigated further. Acknowledgmenb We thank Kim McCracken and Steve Gravina for expert technical assistance, and Holly Stevens

J. Neurochem., Vol. 5 I , No. 6, 1988

and Jeanette Barnhart for preparing the manuscript. This work was supported by a USPHS grant (NS-18150) and by the David S. Ingalls, Sr. Neurological Institute at University Hospitals of Cleveland.

REFERENCES Blasberg R. G., Fenstermacher J. D., and Patlak C. S. (1983) Transport of 2-aminoisobutyric acid across brain capillary and cellular membranes. J. Cereb. Blood Flow Metab. 3, 8-32, Brooks D. J., Gibbs J. S. R., Sharp P., Herold S., Turton D. R., Luthra S. K., Kohner E. M., Bloom S. R., and Jones T. (1986) Regional cerebral glucose transport in insulindependent diabetic patients studied using ["C]3-O-methyl-~-glucoseand positron emission tomography. J. Cereb. Blood Flow Metab. 6,240-244. Cremer J. E. and Seville M. P. (1983) Regional brain blood flow, blood volume and hematocrit values in the adult rat. J. Cereb. Blood Flow Metab. 3, 254-256. Cremer J. E., Cunningham V. J., and Seville M. P. (1983) Relationships between extraction and metabolism of glucose, blood flow, and tissue blood volume regions of rat brain. J. Cereb. Blood Flow Metab. 3, 291-302. DeFronzo R. A., Hendler R., and Christensen N. (1980) Stimulation of counterregulatory hormonal responses in diabetic man by a fall in glucose concentration. Diabetes 29, 125-1 3 1. Duckrow R. B. (1988) Glucose transfer into rat brain during acute and chronic hyperglycemia. Metabol. Brain Dis. (in press). Duckrow R. B. and Bryan R. M. Jr. (1987) Regional cerebral glucose utilization during hyperglycemia. J. Neurochem. 48, 989-993. Duckrow R. B., Beard D. C., and Brennan R. W. (1985) Regional cerebral blood flow decreased during hyperglycemia.Ann. Neurol. 17,267-212. Duckrow R. B., Beard D. C., and Brennan R. W. (1987) Regional cerebral blood flow decreased during chronic and acute hyperglycemia. Stroke 18, 52-58. Ginsberg M. D., Welsh F. A., and Budd W. W. (1980) Deleterious effects of glucose pretreatment on recovery from diffuse cerebral ischemia in the cat. I. Local cerebral blood flow and glucose utilization. Stroke 1I, 347-354. Gjedde A. (1983) Modulation of substrate transport to the brain. Acta Neurol. Scand. 67, 3-25. Gjedde A. and Crone C. (198 I ) Blood-brain glucose transfer: Repression in chronic hyperglycemia. Science 214,456-457. Hank S . I., McCracken K. A., and LaManna J. C. (1986) Regional blood-brain barrier transport of glucose during acute and chronic hyperglycemia in rats. SOC.Neurosci. Abstr. 12, 1261. Hank S. I., Gravina S. A., and Kalaria R. N. (1988) Glucose transporter of the blood-brain bamer and brain in chronic hyperglycemia. J. Neurochem. 51, 1930-1934. Jakobsen J., Sidenius P., Gundersen H. J. G., and Osterby R. (1987) Quantitative changes of cerebral neocortical structure in insulintreated long-term streptozotocin-induced diabetes in rats. Diabetes 36, 591-60 l . Kirk R. A. (1982) Experimental Design: Procedures for Behavioral Sciences. 2nd Ed. pp. 90-1 33. Brookes/Cole, California. LaManna J. C. and Hank S. I. (1985) Regional comparisons ofbrain glucose influx. Brain Res. 326, 299-305. LaManna J. C. and Hank S. I. (1986) Regional studies of bloodbrain bamer transport of glucose and leucine in awake and anesthetized rats. J. Cereb. Blood Flow Metab. 6, 717-723. LaManna J. C. and Hank S. I. (1987) Blood-brain glucose transport and vascular perfusion in acute and chronic hyperglycemia. Can. J. Neurol. Sci. 14, 340. Longstreth W. T. Jr., Diehr P., and Inui D. S. (1983) Prediction of awakening after out-of-hospital cardiac arrest. N. Engl. J. Med. 308, 1378- 1382. McCall A. L., Millington W. R., and Wurtman R. J. (1982) Metabolic fuel and amino acid transport into the brain in experimental diabetes mellitus. Proc. Natl. Acad. Sci. USA 79, 5406-5410. Meyers R. E. and Yamaguchi S. (1977) Nervous system effects of

CBF AND BBB GLUCOSE TRANSPORT IN HYPERGLYCEMIA cardiac arrest in monkeys: Preservation of vision. Arch. Neurol. 34,65-74. Pardridge W. M. and Oldendorf W. H. (1975) Kinetics of bloodbrain barrier transport of hexoses. Biochim. Biophys. Acta 382, 377-392. Pelligrino D. A. and Sharp A. J. (1988) Regional blood-brain barrier (BBB)glucose transport during chronic hypo- and hyperglycemia. FASEB J. 2:A, 1522. Pulsinelli W. A., Waldman S., Rawlinson D., and Plum F. (1982) Moderate hyperglycemia augments ischemic brain damage: A neuropathologic study in the rat. Neurology 32, 1239-1246. Pulsinelli W. A., Levy D. E., Sigsbee B., Scherer P., and Plum F. (1983) Increased damage after ischemic stroke in patients with hyperglycemia with or without established diabetes mellitus. Am. J. Med. 74, 540-544.

1929

Riddle M. C. and Hart J. (1982) Hyperglycemia, recognized and unrecognized, as a risk factor for stroke and transient ischemic attacks. Stroke 13, 356-359. Rubin M. J. and Bohlen H. G. (1985) Cerebral vascular autoregulation of blood flow and tissue p02 in diabetic rats. Am. J. Physiol. 249, H540-H546. Sage J. I., Van Uitert R. L., and Duffy T. E. (1981) Simultaneous measurement of cerebral blood flow and unidirectional movement of substances across the blood-brain bamer: Theory, method and application to leucine. J. Neurochem. 36, 17311738. SiemkowiczE. and Hansen A. J. (1978)Clinical restitution following cerebral ischemia in hypo-, normo- and hyperglycemic rats.Acta Neurol. Scand. 48, 1-8.

J. Neurochem., Yol. 51, No. 6, 1988